U.S. Cotton Subsidies: Drawing a Fine Line on the Degree of Decoupling

Journal of Agricultural and Applied Economics, Apr 2007 by Schmitz, Andrew, Rossi, Frederick, Schmitz, Troy G

Table 3 displays the descriptive statistics of these two models, as well as those of a third model specification referred to as the "Coupled CCP Model". This specification represents an intermediate position of the subsidized supply curve in the interval between the loan rate and the net target price. It differs from the target-linked specification by directly adding only the effective CCP rate to P^sub l^, in order to arrive at P*, the adjusted net target rate (Figure 2). This specification results in simulated free trade prices that are roughly halfway between the free trade prices obtained from the other two models. The average free trade prices from Table 3 confirm this (due to space limitations, we do not show each individual year).

We also modified the basic parameter specification of our models described above by incorporating an alternative set of elasticity estimates. For this "Alternative (PSS)" Model, we utilize elasticity estimates employed by Poonyth et al. (2004) and Sumner (2003) in their studies.11 Table 4 presents the average impacts on world price simulated from this alternative model, and offers a comparison with the results from the previous simulations, including Sumner's price impact as well.12 Again, the comparison period covers the crop years 1999-2000 through 2003-2004. The alternative simulations have somewhat less of an effect on the free trade-world price differential, although the relative variance is higher as observed in the increased coefficient of variation for each specification type (e.g., target-linked, loanlinked). This is so even though the Alternative (PSS) Model employs a larger supply elasticity; the much greater increase in the export elasticity effectively negates the positive effect of these factors, and actually reduces the price distortion as compared to the initial models.

Conclusion

While this paper places focus on the impact of U.S. cotton subsidies, it has applicability to other major U.S. crops covered by U.S. farm subsidy programs. The effects of U.S. agricultural policy are increasingly being scrutinized, especially in view of the successful Brazilian challenge to U.S. cotton subsidies in the WTO. Our results show that the impact of U.S. cotton subsidies can be significant, as also found to be the case in the Brazilian challenge (Powell and Schmitz). The significant price suppression terminology used in the WTO ruling unfortunately does not provide guidance as to where producers respond along the price continuum in our model. Moreover, the findings in the WTO rulings do not give an empirical gauge as to the degree of decoupling. For example, the average negative price impact of 12.4% obtained from our "Loanlinked Model" may be considered significant despite the low degree of decoupling in this specification.

Alternatively, our fully coupled "Targetlinked Model" shows an average negative price impact of 20.5%. The degree to which cotton subsidies are coupled to production is still open to debate, as is true for other commodities supported by U.S. price support programs. In this paper, we present the effects of both a target-linked and a loan-linked specification. As each specification represents an extreme case of the coupling-decoupling debate, we also offer an intermediate case in which CCPs are fully coupled while direct payments are not. The average negative price impact for this intermediate scenario ("Coupled CCP Model") is 16.9%. Within our flexible framework, one can easily model the impact of U.S. cotton subsidies where only a percentage of CCPs are coupled to production.